HAL Id: hal-02600211
https://hal.inrae.fr/hal-02600211
Submitted on 16 May 2020
HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.
Comparison of the performances of Stir Bar Sorptive Extraction (SBSE) coupled to GC-MS-MS and UHPLC-MS-MS for the analysis of pesticides in
freshwaters
C. Guillemain, C. Margoum, A. Assoumani, Marina Coquery
To cite this version:
C. Guillemain, C. Margoum, A. Assoumani, Marina Coquery. Comparison of the performances of Stir Bar Sorptive Extraction (SBSE) coupled to GC-MS-MS and UHPLC-MS-MS for the analysis of pesticides in freshwaters. 8th European Conference on Pesticides and Related Organic Micropollutants in the Environment and 14th Symposium on Chemistry and Fate of Modern Pesticides, Sep 2014, Ioannina, Greece. pp.1, 2014. �hal-02600211�
www.irstea.fr
Comparison of the performances of Stir Bar Sorptive Extraction (SBSE) coupled to GC-MS-MS and UHPLC-MS-MS for the analysis of pesticides in freshwaters
FRESHWATER SYSTEMS, ECOLOGY AND POLLUTIONS RESEARCH UNIT – Irstea, Lyon – Villeurbanne Center - France
C. Guillemain, C. Margoum, A. Assoumani, M. Coquery
Contact : [email protected]
September 18-21, 2014 Ioannina, Greece
RESULTS AND DISCUSSION
METHODS COMPARISON
SBSE-LD-LC-MS-MS method:
21 molecules analyzed
Larger polarity range
> 80% recoveries
Less sensitive method
Non automated method
SBSE-TD-GC-MS-MS method:
More sensitive method
Larger linearity range
> 90% recoveries
Easy and automated method
9 molecules analyzed
CONCLUSIONS
Both methods were validated for the accurate and robust determination of ultra traces of pesticides in freshwaters.
SBSE-LD-UHPLC-MS/MS targets a larger number of pesticides whereas SBSE-TD-GC-MS/MS reaches lower LOQ. Additionally, the automated online thermal desorption system allows faster and easier sample treatment.
The authors thank the French National Research Agency (ANR) through the PoToMAC project (ANR 2011 CESA 022 02) and the French National Agency for Water and Aquatic Environments (ONEMA) for financial support and M. Peyrat for analytical support.
ACKNOWLEDGEMENT
[1] F. David, P. Sandra, J. Chromatogr. A 1152 (1-2), 54 (2007).
[2] C. Margoum, C. Guillemain, X. Yang, M. Coquery, Talanta 116, 1 (2013).
[3] A. Assoumani, C. Margoum, C. Guillemain, M. Coquery, Anal. Bioanal. Chem 406, 2559 (2014).
[4] AFNOR, T90-210 Qualité de l’eau : Protocole d’évaluation initiale des performances d’une méthode dans un laboratoire, 2009, p. 43.
[5] AFNOR, T90-220 Qualité de l'eau : Protocole d'estimation de l'incertitude de mesure associée à un résultat d'analyse pour les méthodes physico-chimiques, 2003, p. 73.
REFERENCES
Analytical uncertainties
Expanded uncertainties for the UHPLC method varied from 9% to 28% depending on the compound and the concentration level, except for SPX at LOQ level.
Similarly, the expanded uncertainties for the GC method varied from 9% to 25%
depending on the compound and the concentration level.
Accuracy and repeatability
Accuracy and repeatability were validated for both methods.
UHPLC method obtains recoveries higher than 94% with RSD below 13% (n = 10 for each level).
GC method reaches recoveries higher than 90%
with RSD below 11% (n = 10 for each level).
Limits of quantification (LOQ)
For the UHPLC method, LOQ are between 5 and 1000 ng/L with recoveries higher than 83% and relative standard deviation (RSD) below 18% (n = 10).
For the GC method, LOQ are between 2.5 and 100 ng/L with recoveries higher than 93% and RSD below 19% (n = 10).
Linearity range
For all compounds, the linearity was validated with a Fisher test or MAD. The ranges for GC method are forty times larger: LOQ to 2000LOQ for GC method and LOQ to 50LOQ for UHPLC method.
Method performances
[4], [5]Limits of quantification (LOQ)
Experimental:
Extraction of 5 duplicates of water spiked at the LOQth (estimated for S/N = k*10).
Experimental LOQ and its standard deviation (sLOQ) were determined.
Evaluation criterion:
LOQ - 2 sLOQ > LOQth - 60% LOQth LOQ + 2 sLOQ < LOQth + 60% LOQth
Accuracy and repeatability
Experimental:
Extraction of 5 duplicates of water spiked at 2 levels: 20%
and 80% of the highest concentration level of the calibration range. Recoveries (R) were determined with calibration curve from extracted standards.
Evaluation criterion:
R - 2 sR > 65%
R + 2 sR < 135%
Linearity range
Experimental:
5 calibration regression curves from the extraction of standards at 6 levels.
2 evaluation criteria:
Fisher test with α = 0.01.
Comparison of the deviations between theoretical and calculated standard concentration values to a maximal acceptable deviation (MAD: from 10% to 60%).
Monitoring of organic pesticides in freshwaters requires developing robust methods that usually involve an extraction step followed by gas or liquid chromatography. Stir Bar Sorptive Extraction (SBSE) is an innovative solvent free sample preparation technique for moderately hydrophobic to hydrophobic organic compounds in aqueous samples
[1]. This fast extraction technique can be followed by two possible desorption modes: liquid desorption (LD) for liquid chromatography or thermal desorption (TD) for gas chromatography. The aim of this work was to validate SBSE-TD-GC-MS-MS and SBSE-LD-UHPLC-MS-MS methods for the determination of several pesticides in surface waters and to compare their performances.
INTRODUCTION
Analytical uncertainties (U)
Experimental:
Extraction of 10 duplicates of 3 natural waters spiked at 3 levels: LQ, 20% and 80% of the highest concentration level of the calibration range.
Determination of uncertainties:
For each concentration level, the average concentration of the samples x and the reproducibility errors srepro were determined.
Then U was calculated as follows, with k = 2:
U (%) = (k × srepro × 100) / x
UHPLC-MS-MS R2 GC-MS-MS R2
AZS 0.02 - 1 0.9995 - -
DMM 0.1 - 5 0.9996 - -
PCM 0.2 - 10 0.9993 0.005 - 10 0.9992
SPX 0.02 - 1 0.9985 - -
TBZ 0.1 - 5 0.9995 - -
ATC 0.05 - 2.5 0.9993 0.005 - 10 0.9986
ATZ 0.005 - 0.25 0.9993 - -
CTU 0.05 - 2.5 0.9994 - -
DFF 0.2 - 10 0.9991 0.05 - 100 0.9981
DIU 1 - 50 0.9992 - -
FMX 0.2 - 10 0.9991 - -
IPU 0.1 - 5 0.9994 - -
MTC 0.005 - 0.25 0.9992 0.005 - 10 0.9992
NFZ 0.2 - 10 0.9991 - -
SMZ 0.05 - 2.5 0.9995 - -
DCA 0.05 - 2.5 0.9984 0.1 - 200 0.9982
DCPMU 1 - 50 0.9992 - -
CFV 0.1 - 5 0.9993 0.025 - 50 0.9972
CPE 0.05 - 2.5 0.9993 0.025 - 50 0.9989 CPM 0.05 - 2.5 0.9992 0.0025 - 5 0.9980 FNT 0.5 - 25 0.9991 0.005 - 10 0.9985 Molecule
Linearity range (µg/L) UHPLC-MS-MS GC-MS-MS
AZS 20 -
DMM 100 -
PCM 200 5
SPX 20 -
TBZ 100 -
ATC 50 5
ATZ 5 -
CTU 50 -
DFF 200 50
DIU 1000 -
FMX 200 -
IPU 100 -
MTC 5 5
NFZ 200 -
SMZ 50 -
DCA 100 100
DCPMU 1000 -
CFV 100 2.5
CPE 50 25
CPM 50 2.5
FNT 500 5
Molecule
LOQ (ng/L)
UHPLC LOQ UHPLC 20% UHPLC 80% GC LOQ GC 20% GC 80%
AZS 16.1 17.1 17.4 - - -
DMM 18.7 16.5 14.6 - - -
PCM 16.8 13.4 14.2 18.6 12.9 15.3
SPX 38.7 26.9 19.4 - - -
TBZ 14.6 11.4 12.7 - - -
ATC 12.5 9.9 11.3 16.9 16.7 20.4
ATZ 13.1 12.9 15.1 - - -
CTU 12.8 16.4 12.9 - - -
DFF 25.0 21.5 15.5 25.0 14.3 16.6
DIU 23.1 14.3 14.4 - - -
FMX 14.8 10.6 10.8 - - -
IPU 21.0 22.2 22.2 - - -
MTC 11.7 16.2 12.1 14.7 14,0 12.6
NFZ 20.7 14.0 16.4 - - -
SMZ 11.3 14.7 15.1 - - -
DCA 16.5 15.6 17.8 18.8 20.2 12.2
DCPMU 19.3 18.3 19.0 - - -
CFV 9.0 10.3 11.6 20.3 12.8 14.4
CPE 16.7 18.8 17.9 18.7 15.5 9.5
CPM 22.4 14.5 10.8 20.8 14.5 11.5
FNT 28.1 21.9 17.1 23.0 14.4 16.1
Molecule
Uncertainties %
F: Fungicide, H: Herbicide, I: Insecticide Log Kow: Octanol-water partition coefficient
MATERIAL AND METHOD
Method synoptic
[2], [3] Sample filtration: Glass fiber filter 0.7 µmExtraction: 20 mL of waters – 3 h at 800 rpm SBSE Twister® l = 20 mm, df = 1.0 mm
Liquid desorption: 15 min in ultrasonic bath, 200 µL methanol/acetonitrile (50/50,v/v)
Thermal desorption: 10 min at 300°C under Helium flow at 75 mL/min
Analysis: UHPLC-ESI(+)-QqQ-MS-MS Shimadzu Nexera 2, API4000 ABSciex, Column HSS T3 Waters® 100 mm*2.1 mm, 1.8 µm
Analysis: GC-Trap-MS-MS
Varian GC3800, Varian MS4000, Column Zebron 5ms Phenomenex® 30 m*0.25 mm, 0.25 µm
Identification / Quantification:
MRM mode acquisition. Confirmation criteria are described in the EU council decision
Internal calibration
Use of labelled molecular analogue as internal standard (IS) and surrogates
Molecule Abbreviation UHPLC-MS-MS GC-MS-MS Use Log Kow
Azoxystrobine AZS X F 2.5
Dimethomorph DMM X F 2.7
Procymidone PCM X X F 3.3
Spiroxamine SPX X F 2.9
Tebuconazole TBZ X F 3.7
Acetochlor ATC X X H 3.0
Atrazine ATZ X H 2.7
Chlortoluron CTU X H 2.5
Diflufenican DFF X X H 4.2
Diuron DIU X H 2.7
Flumioxazine FMX X H 2.6
Isoproturon IPU X H 2.9
Metolachlor MTC X X H 3.1
Norflurazon NFZ X H 2.5
Simazine SMZ X H 2.3
Dichloroaniline DCA X X metab H 2.7
3-(3,4-dichlorophenyl)-
1methylurea DCPMU X metab H 2.5
Chlorfenvinphos CFV X X I 3.8
Chlorpyrifos Ethyl CPE X X I 4.7
Chlorpyrifos Methyl CPM X X I 4.0
Fenitrothion FNT X X I 3.3
